The Sustainable Draw Rates Study
April 19, 2019
Written by Matthew White, CFA®, Investment Analyst
SUSTAINABLE DRAW RATES STUDY: PART TWO
Part two of the Sustainable Draw Rates Study examines the results of different asset mixes over varying time projections.
Results by Decade
After running our simulation, we can see that at ten years, all of our portfolios have virtually no chance of exhausting all funds. This makes intuitive sense as the investor should only have spent 55% of their original portfolio value (5% for ten years but increased by inflation each year). Unless one of the portfolios lost 45% over ten years, some assets should remain. Not much more to learn here, so let’s move to the 20-year test, when portfolios would need to have some sort of positive return to avoid failure.
Projecting out 20 years, we see that the probability of success has dropped below 100% for all portfolios. We can begin to make some distinction in results, though nothing profound. The All Stocks portfolio has an 85% probability of success. This is better than the all bonds portfolio which only has a 70% probability of success. However, we see that a combination of the two, the 50/50 Stocks & Bonds portfolio has a higher probability of success at 93%. There is some rational explanation behind these differing probabilities.
Over 20 years, an investor would have spent 123% of their original portfolio value, so all a portfolio would have to do to succeed is grow by 1% per year. The All Stocks portfolio is more volatile and should have some paths which would be slightly negative to flat over 20 years after recovering from a large downturn. The steadier 50/50 Stocks & Bonds portfolio would have slowly progressed and handily provided the necessary growth to at least fund 20 years of distributions with some residual value.
We see that the multi-asset diversified portfolios have between 93% and 98% probabilities of success, indicating that there may be some value to adding non-traditional investments to portfolios. However, the difference in the probability of success is not significantly different from the 50/50 Stocks & Bonds portfolio and it is hard to derive any hard conclusions yet.
One conclusion that might be drawn, answers our original hypothesis: a portfolio of stocks is not the optimal solution for an investor seeking to spend money and maintain some level of their assets. Before leaping to this inference though, we must consider that investors often have time horizons which are greater than 20 years. Retirees have potential time horizons of 30 years or more, and institutions can have perpetual time horizons. Let’s move forward to our 30-year observation to see what information that provides us.
Going out 30 years, we see a material divergence in outcomes begin to occur. This time period has important implications – for individual investors in particular – as 30 years is a common duration which individuals and their advisors assume for retirement. An all bonds portfolio is all but reasonably expected to run out of assets over 30 years with a measly 6% probability of success. One should not take the 6% probability of success as barely reaching the finish line at the end of life, rather it should be stated as there is a 94% chance an investor will run out of money at some point over 30 years and fail to have distributions to fund their lifestyle or distribute to others.
Interestingly, the All Stocks portfolio has a higher probability of success over 30 years than the 50/50 Stocks & Bonds portfolio, which is an inversion from the 20-year horizon. Again, we turn to what it takes to win over a particular time period. Over 30 years, a portfolio must rely heavily on growth of principle to fund distributions (which total over 205% of original portfolio value when factoring in inflation). Lower returning portfolios cannot fund this growth, especially when a string of low or negative returns cause the corpus to erode which is further exasperated by increasing distributions from the portfolio. Investors with a time horizon of 30 years require portfolios that have high expected returns in order to fund the growth needed for later distributions.
We also note another interesting observation in the difference between the 50/50 Stock & Bond Portfolio and the 50% Stock Diversified Portfolio. The two portfolios have the same stock exposure and thus the only difference is in the 50% non-stock exposure. We see that the use of non-traditional assets improves the probability of success for a 50% stock portfolio from 53% up to 69%. This reinforces what we began to notice in the 20-year horizon: inclusion of non-traditional investments may improve the likelihood of investors obtaining their goals.
Four decades is typically further than investors can envision or project emotionally. However, extending our analysis to 40 years helps us confirm trends we observed at the 30-year horizon, and obtain observations regarding perpetual draws as the probability of success for each portfolio begins to approach a limit.
We observe that the incremental addition of stocks to portfolios increases the probability of success, particularly when we look at the sequence of stock exposure in the diversified portfolios. This would seem to confirm our initial hypothesis: investors are better off with a 100% allocation to stocks – were it not for another observation. The probability of success for an all stocks portfolio is actually diminished from a 65% stock allocation in combination with diversifying investments. We noticed the same variation in the 20-year and 30-year time horizons as well. Even the not-so-careful eye will notice that all of these probabilities of success are not much better than a coin flip of certainty. However, we deliberately chose a strenuous draw rate in order to determine which asset mix provides higher likelihoods of success, whether they are acceptable to a particular investor or not.
Coming in part three: we examine results by historical stress testing and reach our conclusion.